### Ability of covariates to predict treatment status 

### Gender

t_cov_M1 <- lm(wariant ~ M1, 
               data = PBS_data)

### Age 

t_cov_M3_rekod <- lm(wariant ~ M3_rekod, 
               data = PBS_data)

### Level of education

t_cov_M2_rekod <- lm(wariant ~ M2_rekod, 
                     data = PBS_data)

### Size of place of residence 

t_cov_M5 <- lm(wariant ~  M5, 
               data = PBS_data)

### Economic literacy 

t_cov_M6 <- lm(wariant ~ M6, 
               data = PBS_data)

### Size of household

t_cov_M13 <- lm(wariant ~ M13, 
                data = PBS_data)

### Individual income

t_cov_M15 <- lm(wariant ~ M15, 
                data = PBS_data)

### Household income per capita

t_cov_M16_p<- lm(wariant ~ M16_p, 
                data = PBS_data)

### Participation in election 

t_cov_M17 <- lm(wariant ~ M17, 
                data = PBS_data)

### Support for government

t_cov_M20 <- lm(wariant ~ M20, 
                data = PBS_data)

### Employed 

t_cov_M11_1_b <- lm(wariant ~ M11_1_b, 
                    data = PBS_data)
### Self-employed 

t_cov_M12_2_b<- lm(wariant ~ M12_2_b, 
                    data = PBS_data)
### Unemployed

t_cov_M11_3_b <- lm(wariant ~ M11_3_b, 
                    data = PBS_data)
### Student

t_cov_M11_5_b <- lm(wariant ~ M11_5_b, 
                    data = PBS_data)

lmtest::coeftest(t_cov_M11_5_b, df = Inf)
